Hi-Fi Multi-Speaker English TTS (Hi-Fi TTS)

~291.6 h high-quality English multi-speaker TTS from 10 LibriVox speakers (>=17 h each), 44.1 kHz, with Project Gutenberg text.

• Task: tts
• Languages: en
• Hours: 292
• Domain: audiobook
• License: CC BY 4.0
• Homepage: https://www.openslr.org/109/
• Paper: https://arxiv.org/abs/2104.01497

Recommendation

A strong choice for high-fidelity English multi-speaker or single-speaker TTS — clean 44.1 kHz audio with generous per-speaker hours, and CC BY 4.0 / public-domain sourcing safe for commercial use. Prefer it over CSS10 for English. Distinct from the much larger HiFiTTS-2 (~36.7k h).

Getting the data

Downloadable via VoxKitchen (hifitts, source: openslr, size: 38.5 GB):

vkit docker download --tag slim hifitts --root ./data/hifitts

Subsets: default.

NVIDIA release on OpenSLR (SLR109). Do not confuse with HiFiTTS-2.

Suggested processing

A recommended VoxKitchen pipeline ships in the repository at voxkitchen/templates/pipelines/tts-data-prep.yaml — run it with vkit docker run.